Knowledge Distillation For Wireless Edge Learning
This work addresses the challenge of efficient and private edge learning in wireless networks, offering an incremental improvement over existing methods.
The paper tackles the problem of predicting frame errors in congested wireless environments using a new dataset, finding that privacy-preserving federated learning underperforms local training and proposing a framework that combines synthetic minority oversampling and knowledge distillation to achieve better performance and robustness.
In this paper, we propose a framework for predicting frame errors in the collaborative spectrally congested wireless environments of the DARPA Spectrum Collaboration Challenge (SC2) via a recently collected dataset. We employ distributed deep edge learning that is shared among edge nodes and a central cloud. Using this close-to-practice dataset, we find that widely used federated learning approaches, specially those that are privacy preserving, are worse than local training for a wide range of settings. We hence utilize the synthetic minority oversampling technique to maintain privacy via avoiding the transfer of local data to the cloud, and utilize knowledge distillation with an aim to benefit from high cloud computing and storage capabilities. The proposed framework achieves overall better performance than both local and federated training approaches, while being robust against catastrophic failures as well as challenging channel conditions that result in high frame error rates.